Can Computers Perform Workplace Tasks Better Than People?
Digitalization is connecting people, cities, countries and continents, bringing together a majority of the world’s population in ways that vastly increase our individual and collective potential. Yet the same forces have also made the world more volatile, more complex and more uncertain. The rolling processes of automation and the hollowing out of jobs, particularly for routine tasks, have radically altered the nature of work and life. For those with the right knowledge, skills and character qualities this can be liberating and exciting. But for those who are insufficiently prepared, it can mean the scourge of vulnerable and insecure work, and life without prospects. We are living in a digital bazar where anything that is not built for the network age is going to crack under its pressure. Future jobs are likely to pair computer intelligence with the creative, social and emotional skills of human beings. It will then be our capacity for innovation and our awareness that will equip us to harness machines to shape the world. These factors are driving amazing changes in the demand for skills, and the dilemma for educators is that the kind of things that are easiest to teach and easiest to test are precisely the kind of things that are easiest to digitize, automate and outsource.
But what exactly are computers up to? This has been the subject of much speculation. However, new analyses of OECD’s Survey for Adult Skills provide first-of-its-kind insights into current computer capabilities with respect to human skills. It is possible to use the results from this Survey to understand recent changes in skill demand and then assess the computer capabilities that drive skill demand further in the near future. To do this, Dr. Stuart Elliot worked with a group of experts to study current computer capabilities using the test questions from the Survey of Adult Skills. He identified what questions could be answered by current computer techniques and then compared the computer performance with the performance of adults with varying levels of proficiency.
The findings are worrying: two-thirds of workers in OECD countries are using the literacy, numeracy and digital problem-solving skills measured by the Survey of Adult Skills with a proficiency at a level on par with that of computers. Only 13% of workers use these skills on a daily basis with higher proficiency than computers.
Of course, this analysis covers only one set of work skills and therefore provides a limited basis for forecasting how computers will affect employment and skill demand. Different mixes of skills are needed for different work tasks: for some tasks the skills that were measured by the Survey of Adult Skills will be of primary importance and for some tasks they will be required in combination with other skills, such as common sense, expert reasoning, vision, physical movement, or social interaction.
A more comprehensive program to understand how computers will affect employment would need to assess these other skills as well. Additionally, the analysis does not address issues related to the application or diffusion of the computer capabilities in question. Studies of technology adoption find that widespread application of a new production technology often takes one or more decades. In some cases, it fails to occur altogether. The current economy reflects the economic impact of computer techniques developed several decades ago, but not the capabilities unearthed through recent research. Think of the potential and reality of driverless cars, for instance.
Nevertheless, these analyses can provide important conclusions about the implications of future skill demand for education. Over the coming decades, it is likely that there will be strong economic pressure to apply the computer capabilities for skills measured by the Survey across the economy. This is likely to reverse the recent pattern of increasing proportions of workers using low and mid-level literacy skills. Without knowing where new applications will be successful, it is reasonable to conclude that there will be an overall decrease in demand for those workers – the vast majority – whose proficiency in skills that were measured is no better than that of current computer capabilities. This does not mean that these workers will become unemployed; merely that they will become less valuable for many work tasks, making them more vulnerable.
While it remains difficult to assess the immediate implications of these findings for the world of work (since not every job task that computers can manage will be taken on by computers right away,) the results show that we need much better and more systematic intelligence on the capabilities of computers, currently and prospectively, if we want to educate tomorrow’s workers for their future, rather than our past.